-
Notifications
You must be signed in to change notification settings - Fork 217
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Remove metrics endpoints #6283
Remove metrics endpoints #6283
Conversation
Rough measure of the perf impact of logging every ( $ for i in {1..5}; do ctest -VV -R pi_basic_virtual | grep "Average throughput"; done
81: Average throughput: 57120.40 tx/s
81: Average throughput: 57197.73 tx/s
81: Average throughput: 56755.04 tx/s
81: Average throughput: 56493.09 tx/s
81: Average throughput: 56622.10 tx/s
$ wc -l workspace/pi_basic_virtual_0/out
100197 workspace/pi_basic_virtual_0/out Compared to an equivalent build where we don't log: $ for i in {1..5}; do ctest -VV -R pi_basic_virtual | grep "Average throughput"; done
81: Average throughput: 62644.75 tx/s
81: Average throughput: 60552.30 tx/s
81: Average throughput: 61906.70 tx/s
81: Average throughput: 61003.92 tx/s
81: Average throughput: 62806.77 tx/s
$ wc -l workspace/pi_basic_virtual_0/out
181 workspace/pi_basic_virtual_0/out I think this is a significant enough hit that we don't want it by default, and the current implementation (where apps can easily opt in to logging or gathering metrics as they wish) seems fine. So I'm no longer logging in Addendum: The previous "without" numbers were from $ for i in {1..5}; do ctest -VV -R pi_basic_virtual | grep "Average throughput"; done
81: Average throughput: 62528.69 tx/s
81: Average throughput: 63181.41 tx/s
81: Average throughput: 62403.94 tx/s
81: Average throughput: 62519.58 tx/s
81: Average throughput: 62973.88 tx/s |
…metrics_become_logs
Co-authored-by: Amaury Chamayou <[email protected]>
Instead of us maintaining a metrics object, we provide method event hooks so that registries that want to do this can do it themselves, or else just log. In the governance registry, we log every request and its execution time. We also have some custom tracking of
jwt_refresh
, because it's externally triggered and otherwise hard to track.For apps I currently log everything, but I think that's too verbose. Opening as draft to see what the perf impact is. Think the end result will be that apps can log everything, if they want.